Signal processing and preliminary results in the 1988 Monterey Bay Tomography Experiment.

Abstract

Ocean acoustic tomography is particularly suited to observing mesoscale dynamic processes, which may not be adequately observed by more conventional methods. Ships and buoys are limited in their sampling rates by location and /or transit speed while the tomographic signal samples the current and temperature fields all along its path at the speed of sound. Variation in the travel time of the signal occurs due to inhomogeneity in either the sound speed or the current. The ocean's fluctuation can then be estimated from the travel time perturbation using mathematical inverse methods. The 1988 Monterey Bay Tomography Experiment had several specific goals: to test new technology for real-time transmission of tomographic data to shore, to examine the feasibility of doing acoustic tomography in a coastal environment, and to examine the effects of coastal ocean processes such as surface and internal waves and a rough bottom on the tomography signal. This thesis concentrates on signal design using maximal-length sequences, data recording, and a fast algorithm for a datasynchronous digital correlator receiver in this experiment. The new tomographic data recording system has demonstrated its effectiveness. Preliminary results of the data analysis are given, including power spectra for the arrival time perturbation series in the 0.01 to 0.26 Hz (surface wave) frequency band. These spectra correlate well with surface wave spectra obtained from a wave-measuring buoy. Low-pass filtered time series showing perturbations at internal wave frequencies are also presented.